<<

The redisc~very of the Brazilian three ba~ded a,rmadillo and notes on its conservation status - , •~ ;;

Ilmar B. Santos Funda9ilo Biodiversitas, Rua Maria Vaz de Melo, 7L Belo Horizonte, MG 31260-110 l;lrazil.

Gustavo A. B. da Fonseca Depart~mento de Zoologia, Instituto de Ciencias Biologicas, Universidade Federal de Minas Gerais, Av. Antonio Carlos, 6627. Belo Horizonte, MG 31270-110 .

Sonia E. Rigueira Conservation International, Av. Antonio AbraMo Caram, 820/302. Belo Horizonte, MG 31275-000 Brazil.

Ricardo B. Machado Funda9ilo Biodiversitas, Rua MariaVaz de Melo, 7L Belo Horizonte MG 31260-110 Brazil.

Abstract

A recent survey in the northern part of state, the most recent observations were from Coimbrac Brazil, has revealed the presence of Brazilian three Filho and Moojen in 1958 (Coimbra~Filho, 1972), banded Tolypeutes tricinctus, a in the Alto Jaguaribe region (state of ·) and that had not been seen alive by the scientific (state of Bahia). As far back as 1964-68 a community for at least 20 years. The factors that led questionnaire used in the state of Bahia revealed to the decline of the species continue to operate, that T. tricinctus was already extremely rare and three-banded armadillos face an uncertain because of overhunting (Paiva, 1972) .. future. Intensive surveys in the presumed area of distribution of the species are urgently needed so Only six specimens with complete collecting I that a management plan for this endemic Brazilian information are known from the world's museums edentate can be developed. and recent studies throughout its distribution range failed to locate wild populations (Mares et al., Introduction Wetzel, 1981; A Langguth, pers. comm.). To our knowledge, only one individual has been recorded I' The Brazilian three-banded Tolypeutes in zoological collections (Olney, 1980). T. tricinctus is the rarest of the edentates, not having tricinctus is categorized by the IUCN been seen by scientists for at least 20 years. The "indeterminate" (Thornback and Jenkins, 1982). only other species in the is T. matacus, which The Brazilian Institute of Environment and occurs from and the Malo Grosso region of Renewable Resources (IBAMA) lists the Brazilian Brazil to southern (Wetzel, 1981). T. three-banded armadillo as "endangered". tricinctus is confined to north-eastern Brazil, inhabiting the semi-arid (communities of This paper reports on the finding in !988 that T. drought-adapted plants). It is the only endemic tricinctus still lives in the wild in county, Brazilian armadillo and one of two edentates in the northernmost part of the state of Bahia, close restricted to Brazil (the other being the maned sloth to the border of the state of Bradypus torquatus). T. matacus is still relatively (approximately JO•S 39"W) (see Figure 1). common, although heavily hunted in some·parts of its range (Wetzel, 1981). T. tricinctus, on the other Description of locality hand, has not been seen, collected or reported for at least 20 years. Carvalho (1969) observed wild The range of Tolypeutes tricinctus is confined to individuals and also burned shells in the Sao the semi-arid vegetation formations, known Francisco river region of Bahia state. Before this collectively as , which extend inland

Edentata 1(1), Janl131}' 1994 Pagel! poorest diggers of the armadillos and roll up when disturbed, making them very South .America Northeastern Brazil vulnerable to hunters. Although T. matacus id frequently seen and hunted in parts of its range (Wetzel, 1981), it was described as being the second most abundant armadillo species in south­ western Mato Grosso, Brazil (Schaller, 1983),

The first indications of the presence ofT. tricinctus in the area were given by R. B. I Machado and A. Brandt, directors of the Lear's Macaw Conservation Programme under the auspices of WWF -US and ( Funda9iio Biodiversitas. During a sutvey in late 1988 they discovered two burned shells of the species at houses in the I towns of Canudos and Rosario. According to information from local people, these two specimens had been bought at the popular market that setves I,";,,,, I Caatinga region the whole country and takes place every Sunday in the town of Canudos. Machado Figure 1. Map showing the area of Caatinga region in and Brandt took the shells to the Federal Northeastern Brazil and the location of Canudos in the University of Minas Gerais (Belo Bahia state. Horizonte, Brazil), and an expedition was then sent out to locate live individuals. diagonally from the north-east corner of Brazil for 800,000 sq km (Figure 1). The vegetation Although no individuals were seen in the wild, the communities of the region we visited can be broadly authors bought five live adults (three males and two divided into "arboreal caatinga" and "scrub caatinga" females), and took them to the federal University of (Hueck, 1972). The locality surveyed in this study Minas Gerais. The three males, which had multiple dominated by scrub caatinga, characterized by low­ injuries, died upon arrival, and one of the females growing xerophytic woody plants, cacti and gave birth. As is observed with T. matacus (Meritt, terrestrial bromeliads. Average annual 1971), newborn T. tricinctus resemble the adults in precipitation is usually below 500 mm, and the general morphology, possessing hardened , rainfall can be unpredictable (Hueck, 1972; Sick et but with the eyes and ear pinnae closed. Attempts a/., 1987). The area where T. tricinctus was are under way to keep the two sutviving females in observed has been termed " captivity. They are apparently healthy, and at the Ecosystem"(Sick eta/., 1987) and one of the most time of writing have been in captivity for 20 weeks, remote and unpopulated sections of north-eastern being fed on a diet of fruit, mealworms, boiled eggs Brazil. The Raso is characterized by flat terrain cut and a special jelly developed for captive marmosets. by canyons carved by seasonal rivers. The central The individuals purchased at the market were human settlement of this region is the town of reported to be caught, two days before, in the Canudos (county of Canudos). The whole county vicinity of Canudos. According to information from has approximately 13,794 inhabitants, 4,762 of local people, T. tricinctus is relatively common in which live in towns. the market, although the least abundant of the armadillo species sold locally. In addition to the Observations on Tolypeutes tricinctus five Tolypeutes, we found at the same market, -in one single day, ll common armadillos T. tricinctus can be distinguished from T. matacus novemcinctus (sold at US$ 3.00 each), eight six­ in that the former species has five clawed digits on banded armadillos Euphractus sexcinctus (sold at the forefeet, while the latter has only four (Nowak US$ 3.50 each), and one naked-tailed armadillo and Paradiso, 1983). These two species are the tatouay (sold for US$ 3.00). Euphractus

Page 12 Edentata 1(1), Jmmary 1994 is the species preferred by consumers of the region, mostly limited to the production of sisal. It is followed by Dasypus. The cheapest species is probable that a large fraction of the human Tolypeutes (sold for US$ 0.80 each), and population relies on bush meat for its protein apparently the pries are set proportionally to the requirements. Game in the area under study has weight of each species and individual. Individuals been progressively declining, and it is reported that of all ages are sold, but immature. T. tricinctus an area of a few kilometers radius around the town fetch very low prices. Local people say that three­ of Canudos is now almost free of game species. banded armadillos do not thrive in captivity, and Because living conditions are below the poverty therefore cannot be fattened for later consumption. level, and domestic sources of protein are anyway Euphractus, Cabassous and Dasypus, on the other not easily available, the demand for bush meat is hand, are said to be suitable for fattening and thrive not likely to decrease. Education campaigns on a very broad diet. Therefore, even young designed to curb hunting are therefore unlikely to individuals are sold for a fairly good price. succeed in the short term.

According to personal observations and The only protected area in the Canudos region is information by local people, peccaries, Tayassu the Raso da Catarina Ecological Station (997 sq pecari and T. tajacu, brocket deer Mazama km), established in January 1984, which now gouazoubira, lesser anteaters Tamandua belongs to the IBAMA. The reserve protects a fairly tetradactyla, mocos or rock cavies Kerodon large portion of pristine caatinga habitat. A major rupestris and cavies Cavia aperea are also problem with the reserve is its loosely defined frequently sold at the Canudos market. boundaries, and the fact that local people generally ignore its reserve status. T. tricinctus has been Tolypeutes is considered by hunters to be the reported in the Raso da Catarina Ecological species that is mos\ easily caught, and this may Station, but the "reserve" status does not guarantee account for its current rarity. R. Machado has protection; hunters have been seen frequently visited three previous markets at Canudos and had within its boundaries. Since the area has only two reported that the three other armadillo species were permanent guards, law enforcement is highly present, but that To/ypeutes was absent. This may unlikely to improve in the short term. reflect overhunting. Within the distribution range ofT. tricinctus, there Conservation status and prospects for action are other established protected areas that belongs to IBAMA, such as, Aiuaba Ecological Station, The survey described in the present study is National Park of Ubajara, and National Forest of preliminary and a larger project on the Araripe-Apodi (state of Ceara), National Park of conservation and biology of T. tricinctus is being Serra da Capivara and Sete Cidades, and Urucui­ developed. The ease with which T. matacus is Una Ecological Station (state of Piaui), National observed in areas where it is not hunted, associated Park of Chapada Diamantina (state of Bahia), with the fact the previous attempts failed to locate Itabaianas Ecological Station (state of Sergipe), T. tricinctus in the caatinga region, indicate that Serra Negra Ecological Station (state of the species is under severe hunting pressure. Pernambuco), Ecological Station of S.erid6 (state of Anecdotal information obtained locally seems to Rio Grande do Norte). The total area currently suggest that the species was much more common in under protection in the presumed original the past, sharply declining during the past 10 years. geographical range of the Brazilian three-banded Local people gave the impression that three-banded armadillo is approximately 558,433 ha. None of armadillos were once so abundant that armadillo these protected areas has been surveyed for the hunting were trained not to pursue them. species. Nonetheless, the area is important in that it is the only site where T. tricinctus has been described Given these conditions, and anticipating that in the alive during the past 20 years. The low human short term the Brazilian federal and state wildlife population density may account for the persistence agencies will still continue to be financially and of the species, but the threat is certainly increasing. politically weak, we see only a limited number of options that could succeed in improving the The standard of living of Brazilian northeastern conservation status ofT. tricinctus. We have listed populations is among the lowest in Brazil. Long below some of the most obvious suggestions that we and unpredictable droughts make agriculture and feel have a fair probability of success. the keeping of livestock highly risky. Agriculture is

Eden/a/a I (I), January 1994 Page 13 I. At the international level, we suggest that the spixii and the Brazilian three-banded armadillo. JUCN transfers the species from the "indeterminate" category to, at least, "vulnerable". It is our belief, 3. With information on the general biology of T. however, that T. tricinctus merits "endangered" tricinctus, especially diet and reproductive status, for the following reasons : (a) this is the first behaviour, funds should be made available for the l report of live individuals in the least 20 years; (b) a development of captive-breeding programmes. ! number of intensive field studies conducted within Although three-banded armadillos are known to be its distribution range (Mares et a/., 1981; A. keep in captivity, there is no information available Langguth, pers comm.; see also Wetzel, 1981) demonstrating the impossibility of such ·a failed to detect the species; (c) the species is programme. Coimbra-Filho (1972) states that both endemic to the poorest region of Brazil. The T. matacus and T. tricinctus were kept in captivity population reported in this study, in spite of with good results, although no breeding was inhabiting a fairly remote area, is under heavy attempted. The experience of keeping T. tricinctus i hunting pressure, which is likely to be kept constant at the Federal University of Minas Gerais suggests or increase in subsequent years. Evidence (a) and that captive breeding may constitute a viable (b) suggest that most other populations within the method of improving the conservation status of the range of the distribution of the species are likely to species. have disappeared or fallen below critical density. Furthermore, it may be possible that the original Acknowledgments distribution of T. tricinctus was more restricted or less uniform than previously suspected. The Kent Redford stimulated us in writing this note and existence of only a limited number of museum developing a larger project. Ana Maria P. Fonseca specimens with known locality makes the obtained demographic and socioeconomic data on determination of the geographical range of the Canudos. World Wildlife Fund-US financed this species difficult (see Wetzel, 1981 for distribution work. The Funda~ao Biodiversitas and the map); (d) the only reserve in the region (Raso da Department of Zoology of the Federal University of Catarina) lacks any formal protection, and hunters Minas Gerais provided technical and logistical use it freely. Furthermore, previous owners of the support. land now covered by the reserve have not yet been compensated for their properties, which makes the References reserve's future uncertain; (e) the paving of highway BRI16, which is soon to be completed, Carvalho, J. C. M. 1969. Notas de viagem de urn will improve access to the region, thus increasing zoologo a regiao das caatingas e areas hunting pressure on game species (Sick et a/., limitrofes. Bib/. Cult., Vol. 2, Imprensa da 1987). Furthermore, international conservation Universidade Federal do Ceara. organizations should promote fund-raising to Coimbra-Filho, A. F. 1972. Mamiferos amea~ados support field studies, to improve established de extin~ao no Brasil. In: Especies da Fauna reserves and to secure additional protected areas. Brasileira Amear,:adas de Extinr,:iio. Academia Brasileira de Ciencias. Rio de Janeiro, RJ, 2. At the national, state and local levels, efforts Brasil. should be channelled towards providing funding for Hoeck, K. 1972. As F/orestas da America do Sui. the study of the biology of the species in the region. Editora Poligono and Editora Universidade de General surveys should be made in order to search Brasilia, Sao Paulo, SP, Brasil. for other potential populations of T. tricinctus, Mares, M. A., Willig, M. R., Streilein, K. E. and especially within the protected areas already Lacher Jr, T. E. 1981. The of established by JBAMA. The government agencies northeastern Brazil: a preliminary assessment. should try to secure better protection for the Raso Ann. Carn. Nat. His/. 50(4): 81-137. da Catarina Ecological Station, and conduct or fund Meritt Jr, D.A. 1971. The development of the La education and public awareness campaigns to Plata three-banded armadillo Tolypeutes inform local people of the presence and importance matacus at Lincoln Park Zoo, Chicago. of this protected area. With its relatively large area, International Zoo Yearbook 11: 195-196. the Raso da Catarina Ecological Station represents Nowak, R. M. and Paradiso, J. L. 1983. Walker's an important conservation unit within the caatinga Mammals of the World. Vol. L The Johns ecosystem, and may be crucial to the survival of Hopkins University Press, Baltimore. endangered species such as Lear's macaw Olney, P. J. S. 1980. International Zoo Yearbook, Anodorhynchus leari, Spix's macaw Cyanospitta 20, Zoological Society of London.

Page 14 Edentat/11(1), Januatyl994 Paiva, M. P. 1972. Distribui~ao e abundiincia de better understanding of their diet (including alguns mamiferos selvagens no Estado do seasonal variation), and also investigate the Ceara. Rev. Bras. Bioi. 25: 442-450. possible effects of fires on their foraging behavior Schaller, G. B. 1983. Mammals and their biomass (Drumond, !992). Quantitative data were obtained on Brazilian ranch. Arc. Zoo/, S. Paulo 31: 1- on dietary items, time spent attacking each nest, 36. and food availability, and these parameters were Sick, H.,Gonzaga, L. P. and Teixeira, D. M. 1987. compared for anteaters foraging in unburnt and A Arara-azul-de-Lear, Anodorhynchus leari recently burnt areas. Censuses were carried out to Bonaparte, 1856. Rev. Bras. Zoo/. 3: 441-463. investigate possible effects of burning on the Thornback, J. and Jenkins, M. 1982. The JUCN anteater's foraging. The 's diet Red Data Book. International Union showed seasonal variation in prey selection, and for Conservation of Nature and Natural was also evidently determined by different defence Resources, Gland, Switzerland. mechanisms of the (notably So/enopsis, Wetzel, R. M. 1981. Systematics, distribution, Camponotus and Crematogaster) and ecology and conservation of South American (mainly Apicotermitinae, Nasutitermitinae and Edentates. In: Mammalian Biology in South Termitinae) predated. The seasonal effect showed America (M. A. Mares and H. H. Genoways, that although Shaw et a/. ( 1985) recorded a diet of eds.). Vo1.6. Special Publication Series, 88% ants and· 12% termites during two months at Pymatuning Laboratory of Ecology, University the end of the wet season (February and March), at of Pittsburgh. other times of the year (for example, April, August and October) termites can make up the majority of their diet. Attacks on nests and galleries were always of very short duration, very rarely lasting News more than a minute. Short term effects of fires were investigated by censuses and observations of foraging behavior in GIANT ANTEATERS, MYRMECOPHAGA burnt and unburnt areas. They did not avoid TRIDACTYLA: FEEDING BEHAVIOR AND FIRES recently burnt areas, and there was no evident difference in diet nor prey availability from that One of the most important protected areas in the observed for intact . The short term cerrado (bush savanna) of central Brazil is the effects of, at least light, grass fires would appear to Serra da Canastra National Park (71,525 ha), in the be negligible, probably due to the nests acting as west of the state of Minas Gerais. It is an extensive refuges. The anteaters had no problem in finding plateau, largely composed of open grassland, with refuges for themselves during a fire, in swampy altitudes averaging 1,300 m and reaching 1,496 m areas or forest patches. The effect of fires on the above sea level. Perhaps one of the most significant and communities is undoubtedly, features of the park is the community of edentates it however, complex and significant, and one protects, including six species of important factor is the amount of dry biomass (Dasypus novemcinctus, D. septemcinctus, available. One would predict that infrequent burns Cabassous tatouay, C. unicinctus, Euphractus of a larger accumulated biomass may be more sexcinctus, and Priodontes maximus) along with damaging than the frequent and light fires typical the southern tamandua, Tamandua tetradactyla and of most years in the Park. the giant-anteater, Myrmecophaga tridactyla. Field research on the armadillos was carried out by The study formed part of a Master's degree for the Encarna~ao and Carter (Carter; 1983; Carter and course in Ecology, Conservation and Wildlife Encarna~ao, 1983; Encarna~ao, 1986), and Management of the Biological Sciences Institute of population densities and the feeding behavior of the the Federal University of Minas Gerais (UFMG), park's giant anteaters were studied for two months Belo Horizonte. It was supervised by Anthony in 1978 by Shaw, Carter and Machado-Neto (1985). B.Rylands, Zoology Department, UFMG, supported by the Brazilian Institute for the Environment The Park suffers annually from grass fires, and in (Ibama) responsible for the administration of the 1990/91 a study was made of the feeding behavior Park, and financed by Conservation International and diet of the giant anteater during 11 months (CI) - Brazil Program, Belo Horizonte; U.S.Fish (March 1990 to May 1991), in order to obtain a and Wildlife Service, Washington, D.C., Funda~iio Biodiversitas, Belo Horizonte, and the Higher

Eden/ala 1(1), January 1994 Page 15